CPSC 461: Copyright © 2002 Katrin Becker 1998-2002 Last Modified October 20, 2001 09:25 PM
SIGNATURES REVIEW QUESTIONS
- SHORT ANSWER QUESTIONS
- [4 marks] In your own words, what is a Bloom Filter and why might you use one?
- LONG QUESTIONS
- (worth 15 total)
We want to create a signature for a record file (containing 1 million records) to try and avoid having to search the file for records not found.
The signature is 10K-bits long (10,000 bits)
Each record is run through 4 hash functions to specify 4 bits to set:
int H1 (Record* rec);
int H2 (record* rec).
What kind of data structure do we need for the signature? (an array of 10K elements such as bool Sig[10000] is unacceptable; i.e. the bits need to be packed with no wasted space)
The buffer for the signature is
unsigned char BUF[1250];
Given your signature structure, write the algorithm to set the required bits in the signature for one record. Pseudo-code preferred. Each step must be explained.
-
-CPSC 461: Copyright©x 2000 Katrin Becker 1998-2002 Last ModifiedOctober 20, 2001-09:25 PM